Analysis of Wind Forces on a High-Rise Building by RANS-Based Turbulence Models using Computational Fluid Dynamics

Transcription

1 Analysis of Wind Forces on a High-Rise Building by RANS-Based Turbulence Models using Computational Fluid Dynamics D.V.V. Rajkamal 1 and Ch. Raviteja 2 1 (Post graduate student, V.R. Siddhartha Engineering College, Kanuru, Andhra Pradesh ) 2 (Assistant Professor, V.R. Siddhartha Engineering College, Kanuru, Andhra Pradesh) Abstract: Now a day s construction industry is rapidly growing. In order to sustain such development we have to depend on modern tools and one among them is the computer. Wind force is the mo st predominant force acting on any high-raised structures, since wind force is continuous force acting throughout the life of the building it is necessary to analyze any high rise building for these forces. Presently wind forces are being evaluated with th e help of wind tunnel experiment and full-scale tests. There is need of an alternate source for this analysis as the upcoming buildings are most likely to have irregular shapes and plan; as it will become hard to model these as per scale ratio and following all similarities the alternate method of analysis is being developed. Such analysis is Computational Fluid Dynamics (CFD). In the present study a simple high raised building is considered and is evaluated for various turbulent models by using CFD and is compared with the experimental results obtained from Tokyo Polytechnic University "New frontier of Education and Research in Wind Engineering", to find out the suitable turbulent model. The wind pressure coefficients obtained from this analysis using various turbulent models are compared with the wind pressure coefficients obtained from the wind tunnel experiments. It was found that SST k-ꙍ turbulence model & realizable k - turbulence models are better from the results obtained Keywords: k-ɛ turbulent model, Wind pressure coefficients, CFD, Wind tunnel data. I. INTRODUCTION Wind is an important issue on high-raise buildings Wind causes aerodynamic pressures on the surfaces of the buildings. These wind induced aerodynamic pressures also vary randomly w ith time and space. There is a huge demand for high raise buildings in developing countries and developed countries So in order to fulfill that demand we have to construct high raise buildings designed against the wind induced aerodynamic pressures/forces for both and serviceability. This necessitates the calculation of wind loads/forces based on aerodynamic pressure/force coefficients. The magnitudes of these aerodynamic coefficients depend on various parameters, viz. approach wind characteristics including wind direction, geometry and proportion of the building, presences of surrounding buildings, etc These aerodynamic coefficients have been investigated in a number of ways. Wind tunnel investigations on models of buildings is the most preferred method to calculate aerodynamic coefficients Jigar K. Sevalia et.al (2012) attempted to study the effect of different geometric plan configurations like square, circular, Hexagon and Octagon of Tall building having same plan area. It was found that aerodynamic press ure coefficient is maximum in case of square plan shape and it is minimum in case of circular plan shape of tall building. S. Chitra Ganapathi et.al(2013) has investigated aerodynamic forces on 2-d square lattice tower section using two-equation turbulence model S. Chitra Ganapathi et.al(2013) has studied CFD simulations over 2-D Langle section using different Reynolds-Averaged Navier Stokes (RANS) based turbulence models and assess their performance in evaluating drag and lift coefficients for various angles of wind incidence Alexandre Luis Braun et.al (2009) performed aerodynamic and aero elastic analyses on the CAARC standard building model using numerical simulation. LES turbulence model is used in this study. The numerically obtained results were compared with the wind tunnel measurements. and results are compared M.F. Huanga et.al (2011) made a numerical study on full scale tall building using hybrid RANS and Kinematic simu lation. Claudio Mannini et.al (2011) dealt with the threedimensional simulation of the unsteady flow around a stationary 5:1 rectangular cylinder at zero-degree angle of attack. Detached-Eddy Simulation (DES) turbulence model was adopted. Results obtained have shown satisfactory agreement with experimental results. The results obtained with DES for this benchmark test case suggested that this hybrid method is well suited for complex problems of high-reynolds number bluff body aerodynamics. So in order to calculate the response and affect of wind on buildings we have to calculate th e aero dynamic coefficients. Here, we investigate the aero dynamic coefficients using different turbulence models with different angles (0, 45&90) and compare them. Specifically we address the following: Page 28

2 a) First we created model in analysis work bench and created mesh and applied wind loads in different directions b) Second we compared our results with each individually and with the values of Tokyo university c) Third we applied our insights and finally selected that any turbulent models have draw backs but for a basic case like this it is observed that k turbulent model is better II. TURBULENCE MODELLING Turbulence modeling is the computational procedure to solve and analyse the fluid flow introducing some approximations in the governing differential equations so that required solution is obtained approximately consuming feasible computational memory and time. Turbulence modeling is based on the assumption that the real flow field may be substituted by an imaginary field of mathematically defined continuous func tions. The objective of the turbulence modeling is to develop a set of constitutive relations valid for any general turbulent flow problem which yield sufficiently reliable predictions and offer a degree of universality sufficient to justify their usage in terms of computational effort and accuracy. Many turbulence modeling techniques deal with an approximation to the Navier-Stokes equations in form of averaging the different ranges of turbulent eddy scales. For prediction of turbulent flows, the available approaches of turbulence modeling are (i) Direct Numerical Simulation (DNS) (ii) Large Eddy Simu lation (LES) (iii) Reynolds Averaged Navier Stokes solution (RANS). Here we conducted RANSbased turbulence Models mainly K-epsilon models The K-epsilon model is one of the most common turbulence models, although it just doesn't perform well in cases of large adverse pressure gradients.it is a two equation model, which means, it includes two extra transport equations to represent the turbulent properties of the flow. This allows a two equation model to account for history effects like convection and diffusion of turbulent energy. The first transported variable is turbulent kinetic energy, k. The second transported variable in this case is the turbulent dissipation,. It is the variable that determines the scale of the turbulence, whereas the first variable, k, determines the energy in the turbulence. There are two major formulations of K-epsilon models (see References 2 and 3). That of Launder and Sharma is typ ically called the "Standard" K-epsilon Model. The original impetus for the K-epsilon model was to improve the mixing-length model, as well as to find an alternative to algebraically prescribing turbulent length scales in moderate to high complexity flows. As described in Reference 1, the K-epsilon model has been shown to be useful for free-shear layer flows with relatively small pressure gradients. Similarly, for wall-bounded and internal flows, the model gives good results only in cases where mean pressure gradients are small; accuracy has been shown experimentally to be reduced for flows containing large adverse pressure gradients. One might infer then, that the K-epsilon model would be an inappropriate choice for problems such as inlets and compressors. To calculate boundary conditions for these models see turbulence free-stream boundary conditions. Usual K-epsilon models 1. Standard k-epsilon model 2. Standard k- hybrid model 3. Realizable k-epsilon model 4. Realizable k-epsilon hybrid model 5. RNG k-epsilon model 6. RNG k-epsilon hybrid model K-omega models The K-omega model is one of the most commonly used turbulence models. It is a two equation model that means, it includes two extra transport equations to represent the turbulent properties of the flow. This allows a two equation model to account for history effects like convection and diffusion of turbulent energy. The first transported variable is turbulent kinetic energy, k. The second transported variable in this case is the specific dissipation, ꙍ. It is the variable that determines the scale of the turbulence, whereas the first variable, k, determines the energy in the turbulence. To calculate boundary conditions for this model see turbulence free-stream boundary conditions. Usual K-omega models 1. Realizable k-ꙍhybrid turbulence model 2. Standard k-ꙍturbulence model 3. Standard k-ꙍhybrid turbulence model 4. SST k-ꙍ turbulence model 5. SST k-ꙍ hybrid turbulence model Page 29

3 III. NUMERICAL SIMULATION In the present study a 2-D rectangular building model with dimensions of 0.1 m x 0.2 m in plan has been considered for numerical simulation. The wind direction is considered in different angles with the building walls as 0 0,45 0 &90 0. The details of the computational domain for the building have been shown in below for different angles the overall width of the problem domain perpendicular to the flow direction is 2.5 m which is same as that of the width of the wind tunnel. Typical 2-D view of the computation domain for different angles is shown below By using ansys software, the computational domain has been meshed by choosing the sizing option in fluent The velocity inlet boundary was chosen at a distance of 4.8m from the front face of the building. At the inlet boundary, an uniform wind velocity of 10 m/s has been given with a turbulence intensity of 0.2% (i.e. smooth flow with negligible turbulence), which are similar to the values simulated in the wind tunnel. Further, a turbulent viscosity ratio of 10 is generally adopted at the inlet boundary for numerical simulations as it provides better comparison of drag and lift coefficients, with the experimental results The pressure outlet boundary was chosen at a distance of 6.0m from the rear face (leeward face) of the building at which the gauge pressure is set equal to ze ro and the backflow turbulence parameters are set equal to those set at the inlet boundary in order minimize convergence difficulties in case of numerical reversible flow through the outlet boundary during the solution process. The symmetry (free-slip) boundaries were chosen at a distance of 1.2 m from the side faces of the building at which the velocity in the direction normal to the boundary was set equal to zero and also the derivatives of all the flow variables in the direction normal to the boundary were set equal to zero. Wall (no-slip) boundaries were chosen for all the sides of the bluff body and also for the bottom and top boundaries of the computational domain at which the velocities in the flow and normal to flow directions were set equal to zero. Figure1: Computational domain for high raised building section when the angle of wind is 0 0 Figure2: Computational domain for high raised building section when the angle of wind is 90 0 Figure3: Computational domain for high raised building section when the angle of wind is Page 30

4 IV. RESULTS AND DISCUSSION Unsteady numerical simulations have been carried out for a 1:2:5 rectangular building model under uniform wind flow condition using 5 types of turbulence models available in ANSYS FLUENT software which are hybrid and standard: (i) standard k- turbulence model (ii) standard k- hybrid turbulence model (iii) RNG k- turbulence model (iv) RNG k- hybrid turbulence model(v) realizab le k- turbulence model (vi) realizable k-ꙍ hybrid turbulence model (vii) standard k-ꙍturbulence model (viii) standard k-ꙍhybrid turbulence model: (ix) SST k-ꙍ turbulence model (x) SST k-ꙍ hybrid turbulence model. For the evaluation of pressure coefficients, the reference wind velocity is taken as 10 m/s (same as the uniform input wind velocity) Path Line Plots With the help of path lines we can predict the vortex shedding and the wage region Close up view of path line plot near building on horizontal plane for 0,45 and 90 for standard k- turbulence model is shown below Figure4: Close up view of path line plot near building on horizontal plane for 0 0 Figure5: Close up view of path line plot near building on horizontal plane for Page 31

5 Figure6: Close up view of path line plot near building on horizontal plane for 90 0 Comparison of Mean Pressure Coefficient The deviations in the prediction of mean pressure coefficient make it necessary to examine mean pressure coefficient distribution on sections for the thorough understanding of the performance of turbulence models. Since the mean pressure values obtained using SST k-ꙍmodel are observed to compare with the literature values better than those obtained using other models, the corresponding mean pressure coefficient distributions on all sides of the angle section have been considered as reference for assessing the deviations mentioned above. Figures 7 to 9 show the comparison of mean pressure coefficient distributions on the various sides of the building section using different turbulence models for angles of wind incidence 00, 450, 900, 1350 and 3150 or (-450) respectively. Variation of mean pressure coefficient along the curve length of angle section for (0 0 angle of wind incidence) Page 32

6 Variation of mean pressure coefficient along the curve length of angle section for (45 0 angle of wind incidence) Variation of mean pressure coefficient along the curve length of angle section for (90 0 angle of wind incidence) From above figures we can observe the results there is no much variation for ordinary and hybrid models V. SUMMARY AND CONCLUSIONS In the present study, Unsteady numerical simulations have been carried out for a 1:2 rectangular building model under uniform wind flow condition using 10 types of turbulence models available in ansys FLUENT software the numerical simulation of mean pressure coefficients for 2-building under uniform s mooth flow condition have been validated for selective angles of wind incidence. On the overall performance evalu ation of all the Page 33

7 turbulence models considered in the present study, the numerically obtained total pressure coefficients using KE and k-w models compared well with the data base values for all angles of wind incidence. However, the total pressure coefficient values predicted by models are observed to deviate significantly from the literature values. These deviations in the total pressure coefficient predictions are essentially attributed to the limitations in the evaluation of mean suction pressure coefficient values. Further, the predictions of mean lift coefficient using all these turbulence models compared well with the literature and codal values. The limitations of turbulence models in evaluating the suction pressure coefficients in the wake regions are observed to be less pronounced in evaluating mean lift force than in evaluating mean drag coefficient. The values using realizable k- turbulence model and realizable k- hybrid turbulence model are observed are compare well with the experimental value with some less significant difference when the building section is 0 0 & 90 0 and The values using SST k-ꙍturbulence model and SST k-ꙍ hybride turbulence model are observed are compare well with the experimental value with less difference when the building section subjected to wind in 45 0 hence we can study the wind forces on high raised rectangular buildings using CFD simulations for different angles in SST k-ꙍturbulence model & realizable k- turbulence model. VI. REFERENCES [1] S. Chitra Ganapathi, P. Harikrishna and Nagesh R. Iyer. Numerical investigations of aerodynamic forces on 2-d square lattice tower section using two-equation turbulence models. The Eighth Asia-Pacific Conference on Wind Engineering, December 10 14, 2013, Chennai, India [2] S. Chitra Ganapathi, P. Harikrishna and Nagesh R. Iyer. Numerical assessment of aerodynamic forces on 2-d L-angle section using CFD The Eighth Asia-Pacific Conference on Wind Engineering, December 10 14, 2013, Chennai, India [3]. Huang M.F., HLau I.W., Chan C.M., Kwok K.C.S. and G.Li, A hybrid RANS and kinematic simulation of wind load effects on full-scale tall buildings. Journal Wind Eng. Ind. Aerodyn, 99, [4] Shenghong Huanga, Lib Q.S. and Shengli Xua, Numerical evaluation of wind effects on a tall steel building by CFD. Journal of Constructional Steel Research, 63, [5] Shenghong Huanga, Lib Q.S. and Shengli Xua, Numerical evaluation of wind effects on a tall steel building by CFD. Journal of Constructional Steel Research, 63, [6] Alexandre Luis Braun and Armando Miguel Awruch, Aerodynamic and aeroelastic analyses on the CAARC standard tall building model using numerical simulation. Computers and Structures, 87, [7] IS: 875 (Part3)-1987.Code of practice [8] CFD Online.com [9] Aerodynamic Database of High-rise Buildings by the Tokyo Polytechnic University Page 34